
|
EC473 Homepage
Economic Forecasting Seminar at Colby College
Mondays and Wednesdays 1 - 2:15pm in Diamond 342
Professor Michael Donihue
Email address: Michael.Donihue@colby.edu
Office Hours:
Mondays, Tuesdays, Wednesdays & Fridays 11am - noon
Additional times by appointment.
Course Syllabus:ec473syllabusfall07.pdf
(Acrobat
Reader version)
Materials
on Reserve in the Library
| CEO Work In Progress |
Lecture Notes |
Final US model forecasts: cqem1207.wf1 
Maine model as received from the team: ultimate_final_maine_model.wf1 
CEO as of 1-Feb-08: ceo1207.doc 
What I received from M. Poplaski 
final_yield_curve.xls
GovernmentSector-popped.doc
MaineCEOwrite-up.doc
MonetarySector-popped.doc
PresidentialPrediction.doc
PriceandLaborSectorWriteup-popped.doc
Tradedeficitandoilwriteup-popped.doc
Summary table for US outlook: ceo_ussummary.xls
Programs and Data
Data
EViews programs to calculate residuals
For equations without transformed dependent variables: residrun.prg
For equations with log-transformed dependent variables: lresidrun.prg
For equations with first-difference-of-the-log dependents: dlresidrun.prg
For equations with first-differenced dependent variables: dresidrun.prg
Background Information
Threat of Recession Podcast
Portland Fuels the Maine Economy
CQEM Model Development
(1) Using the 2005 version of the CQEM in the EViews file cqem1205.wf1 and documented in the class handout cqem_fall05.pdf, produce a flow chart for your sector of the model.
(2) Update the coefficient estimates for each of the equations in your sector using the data in cqem1007.wf1, estimating each equation through 2006q4 only.
(3) Use these estimates to produce ex post forecasts for each of your equations for the period 2005q2 through 2007q2.
(4) Evaluate each forecast and think about modifications to your sector of the model.
Maine Model Development
Current EViews data file for updating equations: me1007.wf1
Note addition of the Philadelphia Fed's coincident index.
Resources for the Colby Coincident Index of the Maine Economy (CCI)
FRB Philadelphia State Coincident Indexes
T. Crone FRB working paper on constructing state indexes
Stock & Watson paper on coincident indexes
2001 CCI writeup
MS Excel spreadsheet used to calculate the CCI for the 2006 CEO
|
5-Sep Introduction to Forecasting:
ec473lecture01.ppt
10-Sep EDA: ec473lecture02.ppt
19-Sep Exponential Smoothing: ec473lecture03.ppt
26-Sep Holt-Winters: ec473lecture04.ppt
09-Oct Box-Jenkins Identification: ec473lecture05.ppt
25-Oct Autocorrelation & Development of the CQEM Money Demand Eqn: ec473lecture06.ppt
Forecasting Assignments
10-September: Download the EViews workfile noname.wf1 onto your computer. Produce a plot with a fitted line through your series. Any noteworthy features?
12-September: Using the EViews workfile ec473fall07_nsa.wf1 produce a forecast of your series through Dec-08 using the linear trend projection method.
24-September: Use exponential smoothing to estimate different trend values for your series. Why is the forecast produced by exponential smoothing a straight line?
10-October: Box-Jenkins identification, estimation & forecast
15-October: Box-Jenkins forecast using the seasonally adjusted data in ec473fall07_sa.wf1
29-October: Consumption function estimates. Create an EViews workfile, download data from DLX, estimate your consumption function, work on the econometric specification (e.g., check/correct for serial correlation), and complete an ex post forecast for the last 6 quarters of your data.
|
| News Resources |
Colby Quarterly Econometric Model |
Bangor Daily News
BBC News
The Boston Globe
Business Week
CNN News
CNN/Money Financial News
The Economist
Financial Times
Maine Public Radio
MSNBC
National Public Radio
The New York Times
Portland Press Herald
The Wall Street Journal
Washington Post
Waterville Morning Sentinel
|
EViews Workfiles - US Model
Data for this year's version of the CQEM
December 2005
Documentation: cqem_fall05.pdf
December 2003
December 2001
December 1999
December 1997
December 1996
Maine Model
December 2006
December 2005
December 2003
December 2002
November 1999
December 1997
December 1996
|
| Class Resources |
Data Sources |
Programs and Test Statistics
decomp.prg EViews time series decomposition program
recshade.prg EViews shading program
armamodels.prg EViews program providing analytical 'Mug Shots' for identifying ARMA models
DurbinHtest.xls MS Excel worksheet that performs a Durbin h-test
mape.prg Calculates a Mean Absolute Percent Error statistic
errorstats.prg Calculates Mean Error, RMSE, and MAPE statistics
Economic Impact of Hurricane Katrina
Money/CNN 10-Sep-05
New York Times 4-Sep-05
Washington
Post 2-Sep-05
Economist
2-Sep-05
When will Katrina's
impact appear in the data?
Other Current Events
Bernanke's models
Inverted
Yield Curve?
Academic/Government Research
Donihue's Consumption Wealth Effects paper
Donihue's Potential Output/Phillips Curve paper
Does Consumer Sentiment Predict
Regional Consumption?
Trends
and Implications for the Maine Workforce
The Role of Models and Probabilities
in the Monetary Policy Process
How Important Is The Stock Market Effect on Consumption?
A Primer on the Economics and Time Series Econometrics of Wealth Effects (62 pages)
|
US Bureau of Economic Analysis
US Bureau of Labor Statistics
US Census Bureau
Council of Economic Advisers
2005 Economic Report of the President
Economagic.com
Economy.com
Dismal Scientist
FreeLunch.com
Federal Reserve Board of Governors
Boston Fed: New England Economic Indicators
St. Louis Fed: ALFRED
FRASER
FRED
Maine Department of Labor
Maine State Planning Office
OFFSTATS - Official Statistics on the Web
White House Economic Statistics Briefing Room
Yardini's Economic Network
|
Helpful Hints
To check your understanding of the various forecasting and data analysis methods we cover, try to reproduce the charts and numerical results you see in the lecture notes.
Notes on calculating growth rates
Faculty | Course Descriptions | Requirements
Study Abroad | Electronic
Resources |
Colby Home | Search | Directory
Economics Department
5230 Mayflower Hill Drive Waterville, Maine 04901-8852
T: 207.859.5230 F: 207.859.5229 contact
|
 |